404: Are They Always Bad?
If you manage a website long enough, you're going to encounter 404 errors. And the conventional wisdom is pretty simple: 404s are bad. Fix them.
But is that always true?
Not really.
A 404 is simply a web server saying, "There isn't a page at this address." Sometimes that's a problem. Sometimes it's exactly the right answer.
When a 404 really is a problem
Let's say we redesign a website and change:
/services/web-development
to:
/services/website-development
If the old URL has been around for years, Google knows about it. Other websites may link to it. People may have bookmarked it.
We don't want all of those visitors landing on a 404 page.
This is where a 301 redirect belongs. A 301 tells browsers and search engines that the old page has permanently moved to a new location.
Old page → New equivalent page.
Simple.
In fact, we've written before about testing old website URLs before launching a redesign specifically for this reason. A good launch process should account for URLs that are changing, not just the links appearing on the new website.
But here's the thing: not every 404 has a destination
Here's a different scenario.
Another website links to:
yoursite.com/resources/widget-installation-guide
But that page doesn't exist. Maybe it existed 12 years ago. Maybe the other website got the URL wrong. Maybe nobody has any idea where it came from.
A website scan identifies the incoming link and flags the resulting 404.
What should we redirect it to?
Sometimes the answer is: nowhere.
If there is no current page that reasonably replaces that content, a 404 is an accurate response.
Shouldn't we just redirect it to the homepage?
It sounds reasonable. At least the visitor gets somewhere, right?
But a 301 doesn't mean "we couldn't find this, so here's our homepage."
It means, essentially, "the thing you requested has permanently moved here."
If someone follows a link expecting a widget installation guide and lands on the homepage of a 500-page website, we haven't really helped them. More importantly, the homepage isn't the replacement for that missing content.
Search engines understand this distinction too. Redirecting an old URL to an irrelevant destination can be interpreted as a "soft 404." In other words, changing the HTTP response doesn't necessarily convince a search engine that you've provided useful replacement content.
Our rule of thumb
We'd rather make redirects intentional.
Old page has a clear new equivalent? Redirect it.
A URL changed during a website redesign? Absolutely redirect it.
Someone misspelled a URL, but we can clearly tell what page they intended? A redirect is probably helpful.
Old content is gone, but there's a closely related section that genuinely answers the same need? A redirect may make sense.
A third-party website points to a URL that doesn't exist, and there is no meaningful replacement? Let it 404.
That last one is important. A third-party website having a bad link to your website does not necessarily mean there is something wrong with your website.
A good 404 isn't necessarily a dead end
None of this means we want visitors staring at an ugly server error.
A well-designed 404 page can explain that the requested page isn't available and provide useful ways forward: primary navigation, site search, a link to the homepage, or other common destinations.
The important distinction is that the server still correctly reports 404: Not Found.
So, are 404s bad?
Broken links on your own website are bad. Missing redirects after a redesign are bad. Losing valuable inbound traffic because an obvious redirect wasn't created is bad.
But a legitimate request for something that simply doesn't exist?
That's what a 404 is for.